Alexa traffic ranking idiosyncrasy
May 15th, 2007 by Joost Schrier
Alexa has been around for just over 11 years (110 years in internet time). During this time it has evolved from a simple indexing system to a huge database of information about sites worldwide. It takes them 2 months to index 4.5 billion pages from over 16 million websites and they continuously do this. To find out what they need to know, they do not solely rely on links to find and categorize websites like the search engines do; they also use the Alexa Toolbar and other toolbars that use the same technology.
One of the ’services’ that Alexa offers is their traffic ranking. This is basically a website ranking based on the amount of unique visitors to a website over the past 3 months. These visitors are not counted through logfiles or JavaScript counters, but by the Alexa toolbar of the visitor to the website. So, if I visit a website with my Alexa-enabled browser my visit is counted. Bo and Oscar do not use one of these toolbars so their visit is not counted. As a result their information is biased and not overly reliable. The fact that our website is the third most popular website in the Netherlands Antilles according to Alexa is testimony to that.
The image above is part of a screenshot of the list of the most popular websites in the Netherlands Antilles. Our website is in this list because Alexa also lists all websites that are part of the Open Directory Project and dgfmedia.net is in this list. Therefore our website is also listed under the heading of the Netherlands Antilles in Alexa (together with 407 other websites).
According to Alexa the traffic rank of dgfmedia.net is 335,149 today. This means that there are 335,148 websites that are more popular in the internet community than ours. This number is based on 3 months worth of traffic and based on our server logs we can expect that this number will rise because our website has steadily been receiving more traffic.
The Alexa traffic rank of curacao.com (the number one in Alexa’s “Most Popular in the Netherlands Antilles” list) is 292,528. This traffic rank is better than dgfmedia.net’s rank, but not much and I predict that within a month we will have surpassed curacao.com.
So, how does this prove that Alexa’s traffic rank is biased? The reason is that we know for a fact that curacao.com receives about 10 times more traffic/visitors than dgfmedia.net and even more than 10 times more unique visitors.
The reason for this discrepancy has to do with the audience. Our audience is a lot more tech/seo minded and use toolbars to keep an eye on stuff like PR and Alexa traffic rank. Every time they visit our website their visit is logged with Alexa. The people that visit curacao.com are mostly hapless tourists and are not very likely to use an Alexa-enabled toolbar.
Even though it’s pretty cool to be able to say that we will soon have the most popular website of the Netherlands Antilles (according to Alexa) it is a shame that it doesn’t really mean anything. Funny thing: if you want to sell advertising on your website, advertisers usually look at your Alexa rank to calculate what advertising on your website is worth. It would be nice to be able to charge the same prices for advertising here as they charge over at curacao.com :D










June 10th, 2007 at 2:55 pm
[…] Antilles according to Alexa (for an explanation of their biased information see our article on Alexa’s idiosyncrasy). Bookmark […]